This page is part of the web mail archives of SRFI 77 from before July 7th, 2015. The new archives for SRFI 77 contain all messages, not just those from before July 7th, 2015.
> From: bear <bear@xxxxxxxxx> >> On Wed, 23 Nov 2005, Sebastian Egner wrote: >> Mike wrote: >>> From this, and several other messages in the SRFI 77 archives, I'm >>> getting the impression there's some misunderstanding here about the >>> nature of safe/unsafe mode, as alluded to in the document. >> >> Always possible. I understand safe/unsafe mode to mean that there is >> some form of global switch (either compile-time or run-time) that is >> controlling whether certain arithmetic operations raise an exception >> when an error condition happens---or just press on. I understand that >> the 'unsafe' mode allows a compiler to produce code without range >> checking or tag-bit removal, with all sorts of performance benefits. Is >> this what you mean by safe/unsafe mode? > > Actually, I think all he said was that it controls whether error > checking is required or optional - making all the "is an error" > language relating to exact/inexact arithmetic read as "an error is > signalled" when safe mode is on, and leaving it as is (and signalling > at the discretion of the implementation) when it's off. > > You envision a different kind of control, where it's possible to > specify that errors should *not* be detected or signalled, and I > don't think that's what this says. - somewhat paradoxically, it's not clear that halting a program upon encountering what is deemed an error, is more "safe" than a nuisance. - alternatively a typically useful value for potentially erroneous calculations may be returned by default for all functions for which an explicit exception handler has not been defined, in the absence of a global exception handler being explicitly enabled. (in other words, likely useful values are returned by default, or may halt the program's execution if a global exception handler is enabled explicitly; as may be desired during program development, but not likely otherwise)